ALBERT

Description: Background: Studies of the circulatory time of red blood cells (RBCs) or drug delivery vehicles such as liposomes require a radiolabel (e.g., 51Cr), a fluorescent probe or an affinity tag (e.g., biotin). These techniques may be complicated by progressive loss of the probe from the labeled cells, transfer to other cells, and sometimes by the development of an immune response to the label. Long chain dialkylcarbocyanines (Vybrant DiO, DiI and DiD) are bright, stable lipophilic membrane fluorescent labels that can exhibit fluorescence resonance energy transfer (FRET) and are easily observed by flow cytometry using standard filters (FL1, FL2 and FL4 detection channels respectively). The staining is very stable in vivo and is dose-dependent, thus populations of labeled cells may have discrete fluorescence intensities. We recognized that use of these dyes in combination could permit the simultaneous observation of several different populations of cells within the same subject, together with one or more internal controls. To evaluate this technique, we developed procedures suitable for the fluorescent labeling up to 14 discrete populations of human or rabbit RBCs. The properties of the rabbit RBCs were then evaluated in vivo. Methods: Washed human RBCs were suspended at a 10% hematocrit in 20mM HEPES buffer, and added to an equal volume of freshly prepared 1, 4 or 8 μM DiO, DiI, DiD, or mixtures of two labels; 36 different combinations in total. The cells were incubated with the dye for 60 minutes at 37°C, washed twice in HEPES with 0.1% albumin and examined by flow cytometry. Aliquots from 6 to 15 differently-labeled samples were then combined, and the mixtures re-examined. For in vivo studies, the labeling procedure was repeated with autologous rabbit RBCs at dye concentrations of 1.5, 6, and 12 μM. The labeled rabbit RBCs were washed, suspended in saline and reinfused via the ear vein. Small blood samples were drawn daily for 4 days and weekly for 5 weeks, and the number of fluorescently labeled cells of each color remaining was determined by flow cytometry. Results: Of the 36 possible combinations, 14 discrete populations could be clearly distinguished from “dot-plots” of FL1 vs. FL2, FL1 vs. FL4 and FL2 vs. FL4. FRET was observed for all cells stained with two labels. In vivo evaluation of 8 different labeled rabbit RBC populations showed normal survival. A 10–20% reduction in fluorescence intensity was observed over the study period; this reduction did not compromise the clear identification of each discrete population. Rabbits that were repeatedly exposed to Vybrant-labeled autologous RBCs (up to 15 exposures over 3 years) showed no evidence of accelerated clearance, suggesting that these fluorescent labels do not exhibit significant immunogenicity. Conclusion: The use of two long-chain dialkylcarbocyanines in combination is ideal for studies which require multiple cell populations to be followed within the same experiment, and additionally allows for one or more internal controls to correct for any variability of circulation times. The technique is presently being used for studies of RBC senescence in which multiple discrete density populations of RBCs are individually labeled. If these dyes can be shown to have an acceptably low toxicity, the ability to track multiple populations could also be useful for some clinical applications, e.g., to simultaneously evaluate the survival of RBCs from many different donors in patients with multiple alloantibodies, to identify compatible units for transfusion.

Description: Abstract 1259 The ability for hemoglobin S to polymerize causes increased viscosity, decreased RBC deformability and increased RBC aggregation. Chronic transfusion therapy (CTT) in patients with sickle cell disease (SCD) decreases the percentage of hemoglobin S in the bloodstream. The hematocrit to viscosity ratio (HVR) has been used to estimate red cell oxygen transport effectiveness, a property of blood flow. There are undeniable benefits from CTT, however, there are patients on CTT who still suffer from acute crises. Acute hemorheologic changes after transfusion might contribute to these episodes. We hypothesize that viscosity and aggregation will increase post transfusion. Furthermore, we believe that these changes will result in lower oxygen transport effectiveness as measured by HVR. To test this hypothesis we enrolled 26 patients on chronic transfusion therapy in a prospective study to evaluate blood viscosity and aggregation changes with transfusion. We measured both oxygenated and deoxygenated whole blood viscosity at shear rates from 1s−1 to 1000s−1 and RBC aggregation at native hematocrit using a Rheolog viscometer (Rheologics Co). We also obtained pre and post transfusion blood counts, chemistry panels and markers of inflammation and hemolysis. 14 females and 12 males were enrolled with one patient who was excluded due to incomplete data. The ages and reasons for starting transfusion were similar for male and female patients. As expected, transfusion resulted in significant increases of hemoglobin and hematocrit with a concomitant decrease of percent hemoglobin S and reticulocyte count. There was a trend toward decreased platelet count. Male patients had a significantly higher percent hemoglobin S, reticulocyte count, plasma free hemoglobin and platelet count compared to females. Viscosity increased significantly across all shear rates with transfusion and with deoxygenation. There was no sex difference in viscosity. Deoxygenation and transfusion increased the aggregation index. Change in viscosity positively correlated with increased hematocrit and aggregation index. There was no correlation of viscosity change with change in hemoglobin S, markers of inflammation or hemolysis. Deoxygenation lowered HVR at all shear rates. HVR was significantly lower post transfusion at low shear rates of 1s−1, 2s−1 and 5s−1 (Figure 1). At mid to high shear rates there was no difference in HVR, although as the shear rate increased, the HVR curves crossed with apparently increasing improvement in HVR post transfusion at the highest shear rates. The significant decrease in HVR at low shear correlated with increased aggregation index but not increased hematocrit. In this population of patients on CTT, blood viscosity and RBC aggregation increased post transfusion, thus predisposing them to impaired perfusion in low flow vascular beds. The decrease in HVR at low shear rates correlated with RBC aggregation, providing further evidence for post-transfusion risk of impaired perfusion. These results support our hypothesis and indicate potential adverse effects of transfusion on microvascular blood flow.Figure 1Figure 1. Disclosures: Wood: Novartis: Research Funding; Ferrokin Biosciences: Consultancy; Cooleys Anemia Foundation: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding.

Description: Abstract 1518 Poster Board I-541 Introduction Chronic blood transfusions are commonly used as therapy for sickle cell disease (SCD, HbSS) in order to improve oxygen delivery and minimize complications such as stroke in high-risk children. Vaso-occlusive crises can occur in regions of high shear flow (e.g., major cerebral artery occlusions) or regions of low shear flow (e.g., marrow infarct) leading to acute ischemia and, if severe, to necrosis of affected tissues. Transfusion with normal (AA) RBC causes an increase of hematocrit (H) that is complicated by two opposing factors: increased hematocrit (H) causes a linear increase of oxygen carrying capacity and also an exponential increase of blood viscosity (η). As a consequence, the calculated oxygen transport effectiveness, defined as the ratio of H to η (H/η), is a biphasic function of hematocrit: H/η initially increases with H, reaches a maximum at an optimal H value, and then declines with further increases of H. At equal H and shear rate, sickle (SS) blood has significantly higher viscosity than AA and hence part of the strategy for transfusing SCD patients is to reduce η so as to improve H/η. Viscosity studies at high shear rates indicate that an optimum H can be demonstrated for AA-SS RBC mixtures prepared by adding AA RBC to SS blood to simulate transfusion. In marked contrast, low shear rate results for AA-SS mixtures indicate that there is no optimum hematocrit and H/η always decreases with increasing H (Transfusion 46:912-918, 2006). In order to extend these previous in vitro observations to SCD patients, we have measured blood viscosity and hematocrit using whole blood samples acquired prior to and following routine therapeutic transfusion; H/η was calculated over a wide, physiologically relevant shear rate range. Methods All subjects (n= 8, mean age =18.7 years) had homozygous HbSS disease, were crisis-free for 〉 4 weeks, and were enrolled in a chronic transfusion protocol designed to yield 〈 30% HbS and a post-transfusion H of 30-35%. Blood samples were obtained pre- and within 120 hours post-transfusion. A computer-controller tube viscometer was used to determine blood viscosity (37 °C, 40 mm Hg oxygen tension) over a shear rate range of 1 – 1,000 1/s. Results 1) As anticipated, blood viscosity and the degree of non-Newtonian flow behavior increased with H (24.7% pre-transfusion, 34.6% post-transfusion); 2) the change of H/η from pre- to post- transfusion was markedly affected by shear rate (Figure). As indicated, there is a large adverse effect at low shear (i.e., H/η reduced by 20-25% following transfusion), a neutral effect at about 50-100 1/s, and an improved H/η at high shear (Figure). That is, transfusion with AA RBC to obtain a lower percent SS RBC and a higher H actually impairs oxygen transport effectiveness at low shear and is only beneficial at high shear. Conclusions Clinical experience suggests that transfusion regimens aimed a keeping HbS at 30-50% are effective in preventing recurrent strokes in high-risk children. However, our new in vivo transfusion data suggest that at low shear rates, %HbS must be reduced further for H/η to surpass pre-transfusion levels. We interpret these findings as being consistent with our previous data for AA-SS RBC mixtures. They are also consistent with clinical results indicating lack of efficacy for transfusion in low flow areas (e.g., bone marrow during acute crisis) but highly beneficial effects in high flow regions (e.g., cerebral arteries). Our results thus suggest that benefits of transfusion may vary depending on local flow rates (i.e., shear rates) and organ-specific hemodynamics. Disclosures No relevant conflicts of interest to declare.